Mircea Markus wrote:
Galder Zamarreno wrote:
> I'm about to fix https://jira.jboss.org/jira/browse/JBCACHE-1465
> make sure we don't encounter Thread name issues when using Cache
> instances in LOCAL mode but I wanted to raise the following:
> In clustered mode, cluster name allows for each cache instance to be
> distinguished individually but when it comes to LOCAL mode, there's
> not a cache name as such, so if you had N LOCAL caches in a server,
> how would you differentiate from one another when trying to search for
> that instance in your JMX console?
> I think we could do with a cache name in general rather than limiting
> us to a cluster name which by extension is used as the cache name in a
> replicated environment.
I don't see any reason for not having this, just the benefits.
Another approach would be to add another attribute ("name") under the
<exposeJmxStats> element, though I personally like Galder's approach more.
Actually, I like your idea more :)
Having a name in <exposeJmxStats> or similar jmx enabling tag probably
makes more sense than having a generic name. The problem about a generic
name is that then you have two names, name and cluster name and I could
see getting confusing for users, i.e., do I need to set a cache name for
a clustered cache? if so, what does it represent, cache name unique in
cluster or the cache name as it's known cluster wide...etc.
Sr. Software Maintenance Engineer
JBoss, a division of Red Hat